>>13679542They didn't. Just checked the study and I found this.
>A search of the available variables resulted in 5 items measuring one's >own mental illness (e.g., ”Do you have any emotional or mental >disability?”). All of these items were weakly associated with left-wing >political ideology as measured by self-report, with especially high >rates seen for the “extremely liberal” group. These results mostly >held up in regressions that adjusted for age, sex, and race. For the >variable with the most data (n = 11,338), the difference in the mental >illness measure between “extremely liberal” and “extremely >conservative” was 0.39 dThat's all the variables they've adjusted for, at least they could have adjusted for propensity to go to therapy or something but not even that. In fact, this doesn't even take actual diagnosis into account, kek, just a simple question.
